Many of us are already probably familiar with what is known as ‘microfinance’ (see ‘Microcrisis’ by Hanbo Li), but what about social capital? Social capital has been defined by many scholars. To take one famous instance, Putnam in his book ‘Making Democracy Work : Civic Traditions in Modern Italy’ defined social capital as “features of social organizations, such as trust, norms and networks, that can improve the efficiency of society by facilitating coordinated actions”.
So why is this of interest to economists? Most microfinance programs in low income countries emphasize some sort of group participation in forms of group repayment meetings , shared oaths and sometimes joint liability. The Grameen Bank (a microfinance institution originating from Bangladesh) website claims “there is more to the bank than just the balance sheet: it ties lending to a process of social engineering” and it is very common for microfinance institutions to form groups of women, who then must pay the loan back in frequent installments in a group setting. The microfinance institution’s idea is that these group meetings will build social capital. The repeated interactions will increase trust and knowledge within the group consequently resulting in an improvement in default rates. Most importantly however, can we obtain evidence of these effects out in the field?
There have been empirical and experimental studies attempting to capture these peer effects. In particular, Emily Breza (see references) from MIT estimates the causal effects of peer repayment on individuals’ repayment decisions following a crisis (more specifically the Krishna default crisis). During this crisis 100% of borrowers temporarily defaulted on their loans and after which borrowers gradually decided whether to repay or not.
Breza exploits the fact that all the defaults happened simultaneously, and the subsequent consequence that at the time of the mass default, different borrowers would have had different incentives to repay. For example borrowers close to the end of their loan cycle, are on the verge of receiving a renewed loan hence would have had a strong incentive to complete their payment. On the other hand, borrowers who recently received their loan would have less incentive to make the repayment. Now imagine a borrower of the former type, who is in a group where the majority of the borrowers are of the latter type, or vice versa. Using this variation of incentives within the peer group to instrument for the fraction of peers who repay, Breza is able to provide consistent estimates of the effect of peer repayment on individual repayment.
The findings of this empirical survey shows that peer effects do indeed significantly influence individual decisions to repay. More specifically, if the borrower’s peers shift from full default to full repayment, the individual is 10 to 15 percent more likely to repay. This evidence, coupled with the fact that microfinance institutions boast strikingly stellar repayment rates may imply that this social engineering experiment has been quite a success.
The engineering of peer effects and intentionally investing in social capital is not solely the story of low income countries that are subjects of development assistance programs such as microfinance. We in developed countries also experience such peer effects influencing us, consider for example Alcoholics Anonymous. We have similar people grouped together to build trust and knowledge about each other, ultimately in hope of building enough self confidence and willpower to break away from addiction.
Studying the theory behind behavioural peer effects could shed a lot of light in a variety of contexts. Maybe even suggests some improvements to the design of existing peer groups. The success of the microfinance programs in low income countries have been up for debate, however it seems that their engineering of social capital has been quite a success.
Emily , Breza, “Peer Eﬀects and Loan Repayment: Evidence from the Krishna Default Crisis,” Job market paper. http://economics.mit.edu/files/7295.